A method for evaluating the risk level of a gas station

Abstract

The present invention provides a method for evaluating the risk level of a gas station. The method includes: dividing the structure of the gas station to determine the equipment and components that make up the gas station; determining the signal monitoring points of key equipment and obtaining monitoring signals in real time from the sensors installed at the signal monitoring points; estimating the risk level of the gas station based on the monitoring signals and equipment parameters, and predicting the safety situation by establishing a hidden Markov model. By dividing the structure of the gas station and setting key signal monitoring points, quantifying the risk level of the gas station, and predicting the safety situation based on the hidden Markov model, the risk assessment of the gas station is made more accurate and reliable.

Description

Technical Field

[0001] The present invention belongs to the technical field of risk assessment, and particularly relates to a method for evaluating the risk level of a gas station. Background Art

[0002] With the continuous development of the economy and the continuous improvement of people's living standards, the supply demand for industrial and domestic energy has increased rapidly. In recent years, the energy structure has started to implement transformation and development. The original structure with coal and oil as the core energy will be converted and transitioned to cleaner and more environmentally friendly new energy such as hydropower, nuclear energy, and natural gas. Natural gas, this high-quality energy, has received unprecedented attention and development, and its status in China's energy strategy has also been continuously improving. The main production areas of natural gas in China are concentrated in the west, and most of the natural gas resources are imported from abroad, far from the core consumption areas in the southeast coast of the country. Therefore, long-distance natural gas pipelines and gas supply stations have become the infrastructure construction connecting the natural gas production areas and consumption areas. As an important part of the long-distance pipelines between urban and rural areas, gas stations undertake the core tasks of ensuring gas transmission operation and pipeline safety and stability, and play a crucial role.

[0003] Gas stations are the most basic network units in the natural gas transmission system pipeline network. The natural gas pipeline transportation in any city and rural area needs to pass through this checkpoint first. There are a large number of substances with high explosion and combustion risks such as natural gas and hydrocarbons in gas stations, and there are also a large number of gas industrial equipment and facilities. Once natural gas leaks due to equipment failures or other reasons, it will directly cause fires or even explosions in the gas supply station, which will cause heavy losses to gas enterprises and the energy supply of the entire city and the property and personal safety of the people. Therefore, the safety work of gas stations is the top priority in the operation of gas enterprises. Monitoring the operation status of key equipment in gas stations, conducting risk assessment of gas stations, and proposing effective improvement measures based on the risk assessment results are important parts of the safety work of gas stations. Summary of the Invention

[0004] In order to solve the above problems existing in the prior art, the present invention provides a method for evaluating the risk level of a gas station.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions.

[0006] A method for evaluating the risk level of a gas station includes the following steps:

[0007] Divide the structure of the gas station to determine the equipment and components that make up the gas station;

[0008] Determine the signal monitoring points of key equipment, and obtain monitoring signals in real time from the sensors installed at the signal monitoring points;

[0009] Estimate the risk level of the gas station based on the monitoring signals and equipment parameters, and predict the safety situation by establishing a Hidden Markov Model.

[0010] Furthermore, the gas station includes pipelines, valves, pressure regulators, filters, blow-off pipes, and gas-using equipment.

[0011] Furthermore, the signal monitoring points are set at the inlets and outlets of key equipment, important nodes of pipelines, and equally divided points in the pipeline length direction.

[0012] Furthermore, the method for estimating the risk level of the gas station includes:

[0013] Calculate the process processing degree E, which is used to characterize the comprehensive ability of the gas station to process, transform, and handle gas;

[0014] Calculate the layout density F, which is used to characterize the strength of the association and cooperation between one signal monitoring point and other signal monitoring points;

[0015] Calculate the risk level of the gas station based on the process processing degree E and the layout density F.

[0016] Even further, the method for calculating the risk level of the gas station based on E and F includes:

[0017] Set the process processing degree threshold E0 and the layout density threshold F0;

[0018] If E ≤ E0 and F ≤ F0, the risk level is level 1;

[0019] If E > E0 and F ≤ F0, the risk level is level 2;

[0020] If E ≤ E0 and F > F0, the risk level is level 3;

[0021] If E > E0 and F > F0, the risk level is level 4.

[0022] Even further, the calculation formula for the layout density F is:

[0023] F = k1F1 + k2F2 (1)

[0024] In the formula, k1 and k2 are weighting coefficients, 0 < k1 < 1, 0 < k2 < 1, k1 + k2 = 1, F1 is the normalized value of the average distance between on-site equipment, which is equal to the ratio of the average value of the distances between adjacent equipment to the maximum distance, F2 is the equipment layout efficiency index, and the calculation formula for F2 is:

[0025] F2 = S v / S z ×(1 - ε) (2)

[0026] In the formula, S v is the actual floor area of all devices, S z is the area occupied by the device, ε is the equipment complexity coefficient, which is equal to the average value of the complexity coefficients of all devices. The complexity coefficient of a single device is determined by the complexity of the device shape and takes values between 0 and 1. The larger the value, the more complex the device.

[0027] Furthermore, the calculation formula for the process processing degree E is:

[0028] E = E1 × E2 × E3 × E4 (3)

[0029] In the formula, E1 is the normalized value of the number of key devices, E2 is the normalized value of the average value of equipment performance parameters, E3 is the equipment operation status index, and E4 is the synergy coefficient.

[0030] Furthermore, the method for predicting the security situation based on the hidden Markov model includes:

[0031] S1. Establish a hidden Markov model based on the risk level observation sequence O = {O t}: λ = [π, A, B];

[0032] Among them, O t is the observed value of the risk level at the t-th moment, t = 1, 2,..., T, and T is the current moment; the value of the i-th risk level S i is i, i = 1, 2, 3, 4; π is the initial state probability vector, π = {π i}, π i is the probability that the risk level at the first moment is S i ; A is the state transition probability matrix, denoted as A[a ij , a ij is the probability of transitioning from the risk level S i at the (t - 1)-th moment to the risk level S j at the t-th moment; B is the output observation probability matrix, denoted as B[b ij , b ij is the probability that the observed value of the risk level is S i when the risk level is S j ;

[0033] S2. Initialization:

[0034] α1(i) = π i b i (O1) (4)

[0035] In the formula, α1(i) is the probability that the risk level at the first moment is S i , b i (O1) = bij , O1 = S j , i = 1, 2, 3, 4;

[0036] S3. Iterative calculation:

[0037]

[0038] In the formula, α t+1 (i) is the probability that the risk level at the (t + 1)-th moment is S i , b i (O t+1 ) = b ij , O t+1 = S j , i = 1, 2, 3, 4, t = 1, 2,......, T - 1;

[0039] S4. Output the probability of risk in the gas station:

[0040]

[0041] According to formula (5), the probability of each risk level can be predicted, and according to formula (6), the probability of risk in the gas station can be predicted.

[0042] Furthermore, the method further includes: establishing a time series of monitoring signals, decomposing the time series into a trend series, a periodic series, and a residual series, and predicting the monitoring signals at future times.

[0043] Even further, predict the monitoring signals at future times based on the SARIMA model.

[0044] Compared with the prior art, the present invention has the following beneficial effects.

[0045] The present invention divides the structure of the gas station, determines the equipment and components that make up the gas station, determines the signal monitoring points of the key equipment, obtains the monitoring signals in real time from the sensors installed at the signal monitoring points, estimates the risk level of the gas station based on the monitoring signals and equipment parameters, and predicts the safety situation by establishing a hidden Markov model, realizing the assessment of the risk level of the gas station. The present invention divides the structure of the gas station and sets key signal monitoring points, quantifies the risk level of the gas station, and predicts the safety situation based on the hidden Markov model, making the risk assessment of the gas station more accurate and reliable. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 It is a flowchart of a method for assessing the risk level of a gas station according to an embodiment of the present invention.

[0047] Figure 2 It is a schematic structural diagram of a gas station.

[0048] Figure 3 It is a schematic diagram of the mapping relationship between the hidden state and the observed state of the risk level of a gas station.

[0049] Figure 4 It is a schematic diagram of ACF and PACF. Specific implementation manners

[0050] To make the objectives, technical solutions and advantages of the present invention clearer and more understandable, the present invention will be further described below in conjunction with the accompanying drawings and specific implementation manners. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0051] Figure 1 It is a flowchart of a method for evaluating the risk level of a gas station according to an embodiment of the present invention, including the following steps:

[0052] Step 101: Divide the structure of the gas station to determine the equipment and components that make up the gas station;

[0053] Step 102: Determine the signal monitoring points of the key equipment, and obtain the monitoring signals in real time from the sensors installed at the signal monitoring points;

[0054] Step 103: Estimate the risk level of the gas station based on the monitoring signals and equipment parameters, and perform safety situation prediction by establishing a hidden Markov model.

[0055] In this embodiment, step 101 is mainly used to divide the structure of the gas station. Since the state change law of the equipment operation in the gas station is not easy to directly describe, and there are diverse, complex data volumes, data forms and analysis methods in the gas station, it is necessary to first determine the system structure of the gas station to facilitate in-depth analysis. The gas station is divided into the following 5 links:

[0056] Gas transmission link: Gas transmission is the basic function of the gas station, which ensures the safe and efficient transmission of gas from the gas source point to the user end. Among them, the pipeline is the main carrier of gas transmission, which connects the gas source and each use point. Various valves are used to control the flow direction and flow rate of gas to ensure the safety and flexibility of the transmission process. Through the coordinated work of the pipeline and the valve, the gas station can realize the continuous and stable transmission of gas to meet the gas use needs of users.

[0057] Gas start-stop link: The gas start-stop link controls the supply and cut-off of gas, which is crucial for ensuring the safety and stability of the system. Valves play a key role in the start-stop link, controlling the flow of gas through their opening and closing. In case of an emergency, the gas supply can be quickly cut off to prevent the expansion of the accident; under normal circumstances, the gas flow rate is adjusted according to the demand.

[0058] Gas pressure regulation and flow control link: During the transportation and use of gas, pressure and flow adjustments may be required to meet the gas usage needs of different users or adapt to different transportation conditions. The pressure regulator is the key equipment for realizing the changes in gas pressure and flow. The pressure regulator reduces the high-pressure gas in the transmission pipeline to a pressure suitable for downstream users, adjusts the gas flow according to user needs, and can also prevent the gas from flowing back when the downstream pressure is higher than the upstream. Through the adjustment of the pressure regulator, the gas station can ensure the stability and safety of gas under various usage conditions.

[0059] Gas safety treatment link: As a flammable and explosive substance, the safety treatment of gas is an essential part of the gas station. Filters play an important role in gas safety treatment, removing impurities and pollutants in the gas to ensure the cleanliness of the gas. The blow-off pipe is used to discharge gas under specific circumstances to prevent excessive pressure or other safety hazards. Through measures such as filtration and discharge, the gas station can ensure the safe use of gas and reduce the risk of accidents.

[0060] Gas usage link: The ultimate purpose of gas is to meet the gas usage needs of users. Therefore, the usage link is the final link of the gas station system. Through the treatment and guarantee of the previous links, gas can finally be safely and efficiently supplied to the user end to meet the daily life and industrial production needs of users.

[0061] In this embodiment, step 102 is mainly used to determine the signal monitoring points and obtain the monitoring signals. After dividing the structure of the gas station, signal monitoring points are set for the transportation, start-stop, processing, and use of gas, and corresponding sensors are installed at the monitoring points to obtain the monitoring signals from the sensors. The monitoring signals include gas pressure, temperature, and flow rate, etc.

[0062] In this embodiment, step 103 is mainly used to conduct a risk assessment of the gas station. In this embodiment, first, based on the obtained monitoring signals and equipment parameters, the risk level of the gas station is estimated, that is, the risk level is quantified, such as quantified into 1 - 4 risk levels, and a time series of risk levels is established. Then, a hidden Markov model is established for safety situation prediction. For example, the risk level of the gas station and the probability of existing risks are predicted, etc.

[0063] As an optional embodiment, the gas station includes pipelines, valves, pressure regulators, filters, relief pipes, and gas-using equipment.

[0064] This embodiment provides the equipment that constitutes the gas station. According to the five links of the gas station divided above, the equipment belonging to each link is counted, and the equipment included in the gas station is obtained, such as pipelines, valves, pressure regulators, filters, relief pipes, and gas-using equipment, etc. Figure 2 A schematic structural diagram of a gas station is given.

[0065] As an optional embodiment, the signal monitoring points are set at the inlets and outlets of key equipment, important nodes of pipelines, and equally divided points in the pipeline length direction.

[0066] This embodiment provides the setting method of the signal monitoring points. After dividing the structure of the gas station, signal monitoring points are set for the links such as the transportation, start and stop, processing, and use of gas. The setting principles are as follows:

[0067] Set signal monitoring points at the inlets and outlets of key equipment: At each key equipment in the gas station, such as the inlets and outlets of pressure regulators, filters, valves, etc., one signal monitoring point is set respectively. Real-time monitoring and extraction of parameters such as the flow rate and pressure at the inlets and outlets of key equipment are carried out.

[0068] Set signal monitoring points at important nodes of pipelines: Set signal monitoring points at important nodes of gas pipelines, such as branch points, confluence points, and the midpoint or trisection points of the pipeline length, etc. These nodes are important links in the pipeline system, and the gas flow in the pipeline can be monitored by setting signal monitoring points.

[0069] Set signal monitoring points on the load side: Set signal monitoring points on the load side of the gas station, that is, the gas usage points. For large gas users or key facilities, signal monitoring points should be set at their inlets to monitor the gas usage situation and consumption.

[0070] As an optional embodiment, the method for estimating the risk level of a gas station includes:

[0071] Calculate the process processing degree E, which is used to characterize the comprehensive ability of the gas station to process, transform, and handle gas;

[0072] Calculate the layout density F, which is used to characterize the strength of the association and cooperation between one signal monitoring point and other signal monitoring points;

[0073] Calculate the risk level of the gas station based on the process processing degree E and the layout density F.

[0074] This embodiment provides a technical solution for estimating the risk level of a gas station. To achieve a quantitative assessment of the risk of a gas station, this embodiment proposes two indicators, namely, the process processing degree and the layout density. Based on the magnitudes of these two indicators, the risk level of the gas station is calculated. The process processing degree is a comprehensive reflection of the quantity and processing capacity of key equipment in the gas station, which reflects the comprehensive ability of the gas station to process, transform, and handle gas. Specifically, the process processing degree can be quantified by counting the number of key equipment in the gas station, and these key equipment include, but are not limited to, pressure regulators, filters, etc. These equipment not only affect the process processing degree in terms of quantity, but their performance, specifications, and operating status are also important factors for evaluating the process processing degree. When describing the operation and maintenance safety status of a gas station, the layout density refers to the strength of the association and cooperation between an observation node and other nodes. The greater the layout density, the more strongly the node is associated and coordinated with other nodes, and the more likely the node is to affect the normal operation of the entire gas station, thus easily triggering safety risk problems. Therefore, the layout density can reflect the risk status of the gas station.

[0075] As an optional embodiment, the method for calculating the risk level of a gas station based on E and F includes:

[0076] Set a process processing degree threshold E0 and a layout density threshold F0;

[0077] If E ≤ E0 and F ≤ F0, the risk level is level 1;

[0078] If E > E0 and F ≤ F0, the risk level is level 2;

[0079] If E ≤ E0 and F > F0, the risk level is level 3;

[0080] If E > E0 and F > F0, the risk level is level 4.

[0081] This embodiment provides a technical solution for calculating the risk level of a gas station based on E and F. In this embodiment, by respectively setting a process processing degree threshold E0 and a layout density threshold F0, the process processing degree and the layout density are respectively divided into larger values and smaller values (greater than the threshold is a larger value, less than or equal to the threshold is a smaller value), and then the larger and smaller values of the process processing degree are combined with the larger and smaller values of the layout density to obtain 4 risk levels from 1 to 4. For example, when both the process processing degree and the layout density take smaller values, the risk level is the lowest at 1; when both take larger values, the risk level is the highest at 4. The thresholds E0 and F0 generally take the average values of the process processing degree and the layout density respectively.

[0082] Specifically for level 4 risks, a level 1 risk indicates that both the process processing degree and layout density of the gas station are at a relatively low level, and the operation of the station is relatively safe; a level 2 risk means that the gas station has a low layout density but there are many key devices with a high degree of process processing, and its potential risk is higher than that of a level 1 risk; a level 3 risk is that the gas station has a large layout density. Even if the process processing degree is not high, it has a high failure rate, and the operation of the gas station faces greater risks; a level 4 risk is that both indicators are in a high state, which is the highest risk level. Once abnormal operation occurs, it will cause serious consequences and immediate emergency measures need to be taken for intervention. The measures to be taken for each risk level are shown in Table 1.

[0083] Table 1 Measures for Responding to Multi-level Risks

[0084]

[0085]

[0086] As an alternative embodiment, the calculation formula for the layout density F is:

[0087] F = k1F1 + k2F2 (1)

[0088] In the formula, k1 and k2 are weighting coefficients, 0 < k1 < 1, 0 < k2 < 1, k1 + k2 = 1, F1 is the normalized value of the average spacing of on-site devices, which is equal to the ratio of the average value of the spacing between adjacent devices to the maximum spacing, and F2 is the equipment layout efficiency index. The calculation formula for F2 is:

[0089] F2 = S v / S z ×(1 - ε) (2)

[0090] In the formula, S v is the actual floor area of all devices, S z is the area of the region occupied by the devices, and ε is the equipment complexity coefficient, which is equal to the average value of the complexity coefficients of all devices. The complexity coefficient of a single device is determined by the shape complexity of the device and takes a value between 0 and 1. The larger the value, the more complex the device.

[0091] This embodiment provides a technical solution for calculating the layout density F. The layout density is equal to the weighted sum of the average spacing of on-site devices (normalized value) and the equipment layout efficiency index, as shown in formula (1). The equipment layout efficiency is mainly determined by the ratio of the actual floor area of all devices to the area of the region occupied by the devices, and is also negatively correlated with the complexity of the devices. The specific calculation formula is shown in formula (2).

[0092] As an alternative embodiment, the calculation formula for the process processing degree E is:

[0093] E = E1 × E2 × E3 × E4 (3)

[0094] In the formula, E1 is the normalized value of the number of key devices, E2 is the normalized value of the average value of device performance parameters, E3 is the device operation status index, and E4 is the synergy coefficient.

[0095] This embodiment provides a technical solution for calculating the process processing degree E. The process processing degree E is equal to the product of four factors: the normalized value E1 of the number of key devices, the average value of device performance parameters E2, the device operation status index E3, and the synergy coefficient E4. The number of key devices reflects the basic scale and processing capacity of the gas station, and can be quantified according to the actual number of key devices in the gas station. The average value of device performance parameters represents the efficiency of the device in processing gas, and can be calculated by weighted averaging the performance parameters of key devices. The device operation status index takes into account factors such as the stability, failure rate, and maintenance status of the device, and can be calculated based on the actual operation data of the device. For example, it can be represented by the ratio of the normal operation time of the device to the total operation time. The synergy coefficient reflects the degree of cooperation between devices and the overall operation efficiency. This coefficient can be evaluated according to the actual synergy situation between devices, and relies on the experience and judgment of experts to obtain a coefficient between 0 and 1. The higher the coefficient, the better the synergy effect.

[0096] As an optional embodiment, the method for predicting the security situation based on the hidden Markov model includes:

[0097] S1. Establish a hidden Markov model based on the risk level observation sequence O = {O t}: λ = [π, A, B];

[0098] where O t is the observed value of the risk level at the t-th moment, t = 1, 2,..., T, and T is the current moment; the value of the i-th risk level S i is i, i = 1, 2, 3, 4; π is the initial state probability vector, π = {π i}, and π i is the probability that the risk level at the first moment is S i ; A is the state transition probability matrix, denoted as A[a ij , and a ij is the probability of transitioning from the risk level S i at the (t - 1)-th moment to the risk level S j at the t-th moment; B is the output observation probability matrix, denoted as B[b ij , and b ij is the probability that the observed value of the risk level is S i when the risk level is S j ;

[0099] S2. Initialization:

[0100] α1(i) = π i b i (O1) (4)

[0101] Where α1(i) is the probability that the risk level at the first moment is S i of, b i (O1) = b ij , O1 = S j , i = 1, 2, 3, 4;

[0102] S3. Iterative calculation:

[0103]

[0104] Where α t+1 (i) is the probability that the risk level at the (t + 1)-th moment is S i of, b i (O t+1 ) = b ij , O t+1 = S j , i = 1, 2, 3, 4, t = 1, 2,......, T - 1;

[0105] S4. Output the probability of risk existing in the gas station:

[0106]

[0107] According to equation (5), the probability of each risk level can be predicted, and according to equation (6), the probability of risk existing in the gas station can be predicted.

[0108] This embodiment gives a technical solution for safety situation prediction based on the Hidden Markov Model (HMM). First, by using the Baum-Welch algorithm to iterate the risk level observation sequence O = {O1, O2,......, O T} multiple times, the Hidden Markov Model of the gas station state is obtained. The Baum-Welch algorithm is an unsupervised learning algorithm that can estimate the parameters (state transition matrix A) of the Hidden Markov Model (HMM) in the case of only having observation data without corresponding hidden state data. As Figure 3 shown. In Python, the statsmodels library can be used to implement the Baum-Welch algorithm for the training of HMM. Then, based on the estimated HMM model parameters, the forward algorithm is used to calculate the probability α t+1 (i) that the gas station is in each risk level, as shown in equation (5); by summing up the α T (i) of each risk level, the probability of risk existing in the gas station is obtained, as shown in equation (6).

[0109] As an alternative embodiment, the method further includes: establishing a time series of the monitoring signal, decomposing the time series into a trend series, a periodic series, and a residual series, and predicting the monitoring signal at a future time.

[0110] In this embodiment, by establishing a time series of the monitoring signal, the variation law of the monitoring signal over time is analyzed, and based on this, the monitoring signal at a future time is predicted. The time series of the monitoring signal can be obtained according to the timestamps of the collected data of the monitoring signal. In this embodiment, the time series decomposition method STL (Seasonal-Trend decomposition using LOESS) is first used to decompose the time series of the monitoring signal into three parts: a trend series, a seasonal series, and a residual series. Then trend analysis, seasonal analysis, and residual analysis are performed. Trend analysis: shows the long-term change of the data, whether the device performance is gradually improving or declining; Seasonal analysis: displays the periodic change of the data, which helps to identify which months or quarters the device performance is usually better or worse; Residual analysis: is the remaining part of the original data after removing the trend and seasonal components, which helps to identify outliers or random noise. This embodiment combines the trend series and seasonal series obtained by STL decomposition to predict the change of the monitoring signal (device parameters and maintenance data) at a future time by constructing a prediction model.

[0111] As an alternative embodiment, the monitoring signal at a future time is predicted based on the SARIMA (Seasonal Autoregressive Integrated Moving Average) model.

[0112] In this embodiment, the monitoring signal at a future time is predicted by establishing a SARIMA model. The SARIMA model consists of two parts: an autoregressive model (AR) and a moving average model (MA). Among them, the AR model is a function that predicts a variable through its past observed values, and the MA model is a function that predicts a variable through its past prediction errors.

[0113] Specifically, the SARIMA model is established according to the following steps:

[0114] SS1. Determine the seasonal change period s of the time series.

[0115] Plot the time series curve y t (with the horizontal axis being time t and the vertical axis being the value of the time series y t ), and the seasonal change period s can be obtained. For monthly data (data is collected once a month), if the seasonal change repeats once a year, then the seasonal change period s = 12 (months).

[0116] SS2. Obtain the non-seasonal differencing order \(d\) and seasonal differencing order \(D\) through data stationarization processing.

[0117] The statistical properties (such as mean and variance) of a non-stationary time series change over time, which makes it complex to directly analyze the non-stationary time series. Non-stationary time series usually need to be transformed into stationary time series through differencing transformation. The unit root test method can be used to check the stationarity of the time series.

[0118] If the time series is non-stationary, first perform first-order or second-order differencing. For example, for the time series \(y\) t The first-order difference of \(y\): \(\Delta=(1 - L)y\) t \(=y\) t \(-y\) t-1 until the data becomes stationary, and obtain the non-seasonal differencing order \(d\). If the seasonal component is still obvious, perform multiple seasonal differencings on the sequence until the data becomes stationary, and obtain the seasonal differencing order \(D\). The first seasonal difference of \(y\) t is: \(\Delta\) s \(=(1 - L\) s )\(y\) t \(=y\) t \(-y\) t-s .

[0119] SS3. Determine the non-seasonal autoregressive order \(p\), moving average order \(q\), seasonal autoregressive order \(P\), and moving average order \(Q\).

[0120] SS3.1. Plot the ACF graph and PACF graph of the differenced time series.

[0121] ACF is the autocorrelation function, and PACF is the partial autocorrelation function. The abscissa of the ACF graph represents the lag order, and the ordinate represents the correlation coefficient between the corresponding lag sequence and the original sequence. The abscissa of the PACF graph represents the lag order, and the ordinate represents the partial autocorrelation coefficient. As Figure 4 shown.

[0122] SS3.2. Identify \(q\), \(Q\), \(p\), and \(P\) from the ACF graph and PACF graph as follows:

[0123] Autoregressive term (\(p\)): Determine according to the lag value (abscissa value) corresponding to the first significant difference in the correlation coefficients in the PACF graph.

[0124] Moving average term (\(q\)): Determine according to the lag value (abscissa value) corresponding to the first significant difference in the autocorrelation coefficients in the ACF graph.

[0125] Seasonal autoregressive term (\(P\)): Determine by observing the significant lag values (abscissa values) at multiples of the seasonal period in the ACF graph.

[0126] Seasonal moving average term (Q): Determine by observing the significant lag values (abscissa values) at multiples of the seasonal period in the PACF plot.

[0127] SS4. Obtain the value z of the time series at time t through model fitting t :

[0128]

[0129] In the formula, v t is the noise at time t, and A i , B i , Φ i , Θ i are coefficients to be fitted.

[0130] Replace t in formula (7) with t + n, and the value of the time series at any future time can be predicted through recursion, where n = 1, 2,......

[0131] As described above, it is only the specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by those skilled in the art within the technical scope disclosed by the present invention should be covered by the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims

1. A method for assessing the risk level of a gas station, characterized in that: It includes the following steps: Conduct a structural division of the gas station to determine the equipment and components that make up the gas station; Determine the signal monitoring points of key equipment, and obtain monitoring signals in real time from the sensors installed at the signal monitoring points; the signal monitoring points are set at the inlets and outlets of key equipment, important nodes of pipelines, and equally divided points in the pipeline length direction; Estimate the risk level of the gas station based on the monitoring signals and equipment parameters, and conduct safety situation prediction by establishing a Hidden Markov Model; The method for estimating the risk level of the gas station includes: Calculate the process processing degree E, which is used to characterize the comprehensive ability of the gas station to process, transform, and handle gas; Calculate the layout density F, which is used to characterize the strength of the association and cooperation between one signal monitoring point and other signal monitoring points; Calculate the risk level of the gas station based on the process processing degree E and the layout density F; The calculation formula for the layout density F is: F = k1F1 + k2F2 (1) In the formula, k1 and k2 are weighting coefficients, 0 < k1 < 1, 0 < k2 < 1, k1 + k2 = 1, F1 is the normalized value of the average spacing of on-site equipment, which is equal to the ratio of the average value of the adjacent equipment spacing to the maximum spacing, F2 is the equipment layout efficiency index, and the calculation formula for F2 is: F2 = S v / S z ×(1 - ε) (2) Where S v is the actual floor space occupied by all equipment, S z is the area occupied by the equipment, ε is the equipment complexity coefficient, which is equal to the mean of the complexity coefficients of all equipment. The complexity coefficient of a single equipment is determined by the complexity of the equipment shape and takes a value between 0 and 1. The larger the value, the more complex the equipment.

2. The gas station risk level assessment method according to claim 1, characterized in that: The gas station includes pipelines, valves, pressure regulators, filters, blow-off pipes, and gas-using equipment.

3. The gas station risk level assessment method according to claim 1, characterized in that: The method for calculating the risk level of the gas station based on E and F includes: Set the process processing degree threshold E0 and the layout density threshold F0; If E ≤ E0 and F ≤ F0, the risk level is level 1; If E > E0 and F ≤ F0, the risk level is level 2; If E ≤ E0 and F > F0, the risk level is level 3; If E > E0 and F > F0, the risk level is level 4.

4. The risk level assessment method for gas stations according to claim 1, characterized in that The calculation formula for the process processing degree E is: E = E1 × E2 × E3 × E4 (3) In the formula, E1 is the normalized value of the number of key equipment, E2 is the normalized value of the average value of equipment performance parameters, E3 is the equipment operation status index, and E4 is the synergy coefficient.

5. The gas station risk level assessment method according to claim 3, characterized in that: The method for conducting safety situation prediction based on the Hidden Markov Model includes: S1. Based on the risk level observation sequence O = {O t}, establish a hidden Markov model: λ = [π, A, B]; Among them, O t is the observed value of the risk level at the t-th moment, where t = 1, 2,......, T, and T is the current moment; the value of the i-th risk level S i is i, where i = 1, 2, 3, 4; π is the initial state probability vector, π = {π i}, and π i is the probability that the risk level at the 1st moment is S i ; A is the state transition probability matrix, denoted as A[a ij , and a ij is the probability of transitioning from the risk level S i at the (t - 1)-th moment to the risk level S j at the t-th moment; B is the output observation probability matrix, denoted as B[b ij , and b ij is the probability that the observed value of the risk level is S i when the risk level is S j . S2. Initialization: α1(i)=π i b i (O1) (4) where α1(i) is the probability that the risk level at the first moment is S i , b i (O1) = b ij , O1 = S j , i = 1, 2, 3, 4; S3. Iterative calculation: where α t+1 (i) is the probability that the risk level at the (t + 1)-th moment is S i , b i (O t+1 ) = b ij , O t+1 = S j , i = 1, 2, 3, 4, t = 1, 2,......, T - 1; S4. Output the probability of risks existing in the gas station: According to formula (5), the probability of each risk level can be predicted, and according to formula (6), the probability of risks existing in the gas station can be predicted.

6. The risk level assessment method for gas stations according to claim 1, characterized in that The method further includes: establishing a time series of monitoring signals, decomposing the time series into a trend series, a periodic series, and a residual series, and predicting the monitoring signals at future times.

7. The risk level assessment method for a gas station according to claim 1, characterized in that Predict the monitoring signals at future times based on the SARIMA model.

Images